A New York newspaper mentioned Frenchmen who “content themselves with sipping _thimbles full_ of absinthe.” The reader should have known that the men do not use thimbles for the purpose of drinking, and that _thimblefuls_ are what they sip.
When the proof-reader had a paragraph saying that “the arrivals at the hotels show a falling off of over 100 per cent.,” he should have known that this is an impossibility, since it leaves the arrivals less than none.
When another reader saw something about “the buildings _comprising_ the old brick row,” he should have corrected it to _composing_. Buildings compose the row, and the row comprises buildings.
It would not be fair to expect every proof-reader to be thoroughly up in zoölogical nomenclature. No reader, though, should pass a word like _depuvans_ unchallenged, because that is the best he can make of what is written. He should ascertain in some way that the word is _dipnoans_, or query it for some one else to correct. On the “Century Dictionary” the editor struck out a quotation, “The miracles which they saw, grew by their frequency familiar unto them.” His pencil happened to cross only one word in the first line, and the next proof sent to the editorial room contained the passage, “The miracles which they grew by their frequency familiar unto them.”
These are a few instances of remissness on the part of readers, the last one showing absurdity that should be impossible.
Some things are commonly expected of proof-readers that they can not with any reason be asked to do. When a person whose initials are J. J., for instance, writes them I. I., it is not reasonable to expect them to be printed J. J. A script I is one thing and a J is another; and no one can possibly know that the one which is written is not the right one when there is no clue, as there would be in Iohn. One lesson that writers seem bound not to learn is that proper names should be written plainly. When not written plainly they are very likely to be printed wrong.
Some kinds of changes proof-readers should not make, even if they think the writing is wrong. When a plainly written manuscript, showing care at all points, contains something about the “setting up of the first printing-press,” this should not be printed “setting-up of the first printing press”; neither should _some one_ be changed to _someone_, though the barbarous _someone_ happens to be the “style of the office.” There is no good reason for making a compound of _setting up_, and there is no reason for making anything but a compound of _printing-press_; and _someone_ should certainly be removed from the “style of the office” and the correct _some one_ substituted. These two examples are selected because they were convenient, not for criticism merely, but to enforce the fact that, at least in a book or any work not containing matter from various writers, carefully written manuscript should be followed in every respect. Some authors have in this matter a just cause of complaint against printers; but it is really the result of carelessness on the part of authors in not writing as their matter should be printed and insisting upon having what they want.